AdamGleave

• AdamGleave 17 Mar 2024 0:55 UTC

  6 points

  0 ∶ 0

  in reply to: RyanCarey’s comment on: More peo­ple get­ting into AI safety should do a PhD

  This is an important point. There’s a huge demand for research leads in general, but the people hiring & funding often have pretty narrow interests. If your agenda is legibly exciting to them, then you’re in a great position. Otherwise, there can be very little support for more exploratory work. And I want to emphasize the legible part here: you can do something that’s great & would be exciting to people if they understood it, but novel research is often time-consuming to understand, and these are time-constrained people who will not want to invest that time unless they have a strong signal it’s promising.

  A lot of this problem is downstream of very limited grantmaker time in AI safety. I expect this to improve in the near future, but not enough to fully solve the problem.

  I do like the idea of a more research agenda agnostic research organization. I’m striving to have FAR be more open-minded, but we can’t support everything so are still pretty opinionated to prioritize agendas that we’re most excited by & which are a good fit for our research style (engineering-intensive empirical work). I’d like to see another org in this space set-up to support a broader range of agendas, and am happy to advise people who’d like to set something like this up.

More peo­ple get­ting into AI safety should do a PhD

2023 Align­ment Re­search Up­dates from FAR AI

What’s new at FAR AI

• AdamGleave 16 Jun 2023 17:11 UTC

  40 points

  13 ∶ 0

  on: Cri­tiques of non-ex­is­tent AI safety labs: Yours

  As someone who did recently set up an AI safety lab, success rates have certainly been on my mind. It’s certainly challenging, but I think the reference class we’re in might be better than it seems at first.

  I think a big part of what makes succeeding as a for-profit tech start-up challenging is that so many other talented individuals are chasing the same, good ideas. For every Amazon there are 1000s of failed e-commerce start-ups. Clearly, Amazon did something much better than the competition. But what if Amazon didn’t exist? What if there was a company that was a little more expensive, and had longer shipping times? I’d wager that company would still be highly successful.

  Far fewer people are working on AI safety. That’s a bad thing, but it does at least mean that there’s more low-hanging fruit to be tapped. I agree with [Adam Binks](https://​​forum.effectivealtruism.org/​​posts/​​PJLx7CwB4mtaDgmFc/​​critiques-of-non-existent-ai-safety-labs-yours?commentId=eLarcd8no5iKqFaNQ) that academic labs might be a better reference class. But even there, AI safety has had far less attention paid to it than e.g. developing treatments for cancer or unifying quantum mechanics and general relativity.

  So overall it’s far from clear to me that it’s harder to make progress on AI safety than solve outstanding challenge problems in academia, or in trying to make a $1 bn+ company.

• AdamGleave 12 Jun 2023 6:31 UTC

  31 points

  6 ∶ 0

  on: Why I Spoke to TIME Magaz­ine, and My Ex­pe­rience as a Fe­male AI Re­searcher in Sili­con Valley with Sex­ual Harass­ment/​Abuse

  Thanks Lucretia for sharing your experience. This cannot have been an easy topic to write about, and I’m deeply sorry these events happened to you. I really appreciated the clarity of the post and found the additional context above the TIME article to be extremely valuable.

  I liked your suggestions for people to take on an individual level. Building on the idea of back channel references, I’m wondering if there’s value in having a centralised place to collect and aggregate potential red flags? Personal networks can only go so far, and it’s often useful to distinguish between isolated instances and repeated patterns of behaviour. The CEA CH team partially serves this role within EA, but there’s no equivalent in the broader Silicon Valley or AI communities.

AI Safety in a World of Vuln­er­a­ble Ma­chine Learn­ing Systems

• AdamGleave 6 Dec 2022 21:56 UTC

  51 points

  14 ∶ 0

  in reply to: lastmistborn’s comment on: Why did CEA buy Wytham Abbey?

  For people not familiar with the UK, the London metropolitan area houses 20% of the UK’s population, and a disproportionate share of the economic and research activity. The London-Cambridge-Oxford triangle in particular is by far the research powerhouse of the country, although there are of course some good universities elsewhere (e.g. Durham, St Andrews in the north). Unfortunately, anywhere within an hour’s travel of London is going to be expensive. Although I’m sure you can find somewhat cheaper options than Oxford, I expect the cost savings will be modest (noting Oxford is already cheaper than central London), and you’ll likely lose something else (e.g. location is harder to get to, or is some grungy commuter town).

  I would like to hear if CEA considered non-Oxford locations (as there’s an obvious natural bias given CEA is headquartered in Oxford), but it wouldn’t surprise me if the benefit of CEA staff (who will often be running the events) having easy access to the venue genuinely outweighed any likely modest cost savings from locating elsewhere.

• AdamGleave 6 Dec 2022 21:47 UTC

  66 points

  24 ∶ 1

  in reply to: MatthewDahlhausen’s comment on: Why did CEA buy Wytham Abbey?

  A 30 person office could not house the people attending, so you’d need to add costs of a hotel/​AirBnB/​renting nearby houses if going down that option. Even taking into account that commercial rest estate is usually more expensive than residential, I’d expect the attendee accommodation cost to be greater than the office rental simply because people need more living space than they do conference space.

  Additionally in my experience retreats tend to go much better if everyone is on-site in one location: it encourages more spontaneous interaction outside of the scheduled time. There are also benefits to being outside a city center (too easy for people to get distracted and wander off otherwise).

  Was Wytham a wise investment? I’m not sure, I’d love to see a calculation on it, and it probably comes down to things like the eventual utilization rate. But I think a fairer reference class would be “renting a conference center plus hotel” than “renting a 30-person office”.

• AdamGleave 14 Nov 2022 3:47 UTC

  2 points

  0 ∶ 0

  in reply to: spencerg’s comment on: Thoughts on le­gal con­cerns sur­round­ing the FTX situation

  Note I don’t see any results for FTX Foundation or FTX Philanthropy at https://​​apps.irs.gov/​​app/​​eos/​​determinationLettersSearch So it’s possible it’s not a 501(c)(3) (although it could still be a non-profit corporation).

• AdamGleave 9 Nov 2022 2:48 UTC

  23 points

  3 ∶ 0

  in reply to: Sabs’s comment on: FTX will likely go bankrupt. What we know and some fore­casts on what will hap­pen next

  Disclaimer: I do not work for FTX, and am basing this answer off publicly available information, which I have not vetted in detail.

  Nick Beckstead in the Future Fund launch post described several entities (FTX Foundation Inc, DAFs) that funds will be disbursed out of: https://​​forum.effectivealtruism.org/​​posts/​​2mx6xrDrwiEKzfgks/​​announcing-the-future-fund-1?commentId=qtJ7KviYxWiZPubtY I would expect these entities to be sufficiently capitalized to provide continuity of operations, although presumably it’ll have a major impact on their long-run scale.

  IANAL but I’d expect the funds in the foundation/​DAF to be fairly secure against bankruptcy or court proceedings. Bankruptcy courts can’t just claw back money arbitrarily from other creditors, and limited liability corporations provide significant protection for directors. However, I’d expect assets donated to FTX Foundation or associated DAFs to largely be held in-kind (again, this is speculation, but it’s standard practice for large philanthropic foundations) not liquidated for cash. These assets mark-to-market value are likely worth a lot less than they were a week ago.

• AdamGleave 19 Aug 2022 22:41 UTC

  7 points

  0 ∶ 0

  in reply to: aaronhamlin’s comment on: Long-Term Fu­ture Fund: De­cem­ber 2021 grant recommendations

  Hi Aaron, thanks for highlighting this. We inadvertently published an older version of the write-up before your feedback—this has been corrected now. However, there are still a number of areas in the revised version which I expect you’ll still take issue with, so I wanted to share a bit of perspective on this. I think it’s excellent you brought up this disagreement in a comment, and would encourage people to form their own opinion.

  First, for a bit of context, my grant write-ups are meant to accurately reflect my thought process, including any reservations I have about a grant. They’re not meant to present all possible perspectives—I certainly hope that donors use other data points when making their decisions, including of course CES’s own fundraising materials.

  My understanding is you have two main disagreements with the write-up: that I understate CES’s ability to have an impact on the federal level, and that the cost effectiveness is lower than you believe to be true.

  On the federal level, my updated write-up acknowledges that “CES may be able to have influence at the federal level by changing state-level voting rules on how senators and representatives are elected. This is not something they have accomplished yet, but would be a fairly natural extension of the work they have done so far.” However, I remain skeptical regarding the Presidential general for the reasons stated: it’ll remain effectively a two-candidate race until a majority of electoral college votes can be won by approval voting. I do not believe you ever addressed that concern.

  Regarding the cost effectiveness, I believe your core concern was that we included your total budget as a cost, whereas much of your spending is allocated towards longer-term initiatives that do not directly win a present-day approval voting campaign. This was intended as a rough metric—a more careful analysis would be needed to pinpoint the cost effectiveness. However, I’m not sure that such an analysis would necessarily give a more favorable figure. You presumably went after jurisdictions where winning approval voting reform is unusually easy; so we might well expect your cost per vote to increase in future. If you do have any internal analysis to share on that then I’m sure I and others would be interested to see it.

Long-Term Fu­ture Fund: De­cem­ber 2021 grant recommendations

• AdamGleave 6 Jul 2022 19:13 UTC

  2 points

  0 ∶ 0

  on: Who’s hiring? (May-Septem­ber 2022)

  The Fund for Alignment Research is hiring for an Operations Manager, ML Research Engineers and Communications Specialists (technical writing, graphics design, etc)

In­tro­duc­ing the Fund for Align­ment Re­search (We’re Hiring!)

• AdamGleave 5 Jul 2022 1:35 UTC

  11 points

  0 ∶ 0

  in reply to: Guy Raveh’s comment on: Why AGI Timeline Re­search/​Dis­course Might Be Overrated

  You could argue from a “flash of insight” and scientific paradigm shifts generally giving rise to sudden progress. We certainly know contemporary techniques are vastly less sample and compute efficient than the human brain—so there does exist some learning algorithm much better than what we have today. Moreover there probably exists some learning algorithm that would give rise to AGI on contemporary (albeit expensive) hardware. For example, ACX notes there’s a supercomputer than can do $10^17$ FLOPS vs the estimated $10^16 needed for a human brain. These kinds of comparisons are always a bit apples to oranges, but it does seem like compute is probably not the bottleneck (or won’t be in 10 years) for a maximally-efficient algorithm.

  The nub of course is whether such an algorithm is plausibly reachable by human flash of insight (and not via e.g. detailed empirical study and refinement of a less efficient but working AGI). It’s hard to rule out. How simple/​universal we think the algorithm the human brain implements is one piece of evidence here—the more complex and laden with inductive bias (e.g. innate behavior), the less likely we are to come up with it. But even if the human brain is a Rube Goldberg machine, perhaps there does exist some more straightforward algorithm evolution did not happen upon.

  Personally I’d put little weight on this. I have <10% probability on AGI in next 10 years, and think I put no more than 15% on AGI being developed ever by something that looks like a sudden insight than more continuous progress. Notably even if such an insight does happen soon, I’d expect it to take at least 3-5 years for it to gain recognition and be sufficiently scaled up to work. I do think it’s probable enough for us to actively keep an eye out for promising new ideas that could lead to AGI so we can be ahead of the game. I think it’s good for example that a lot of people working on AI safety were working on language models “before it was cool” (I was not one of these people), for example, although we’ve maybe now piled too much into that area.

• AdamGleave 4 Jul 2022 2:26 UTC

  43 points

  0 ∶ 0

  on: Why AGI Timeline Re­search/​Dis­course Might Be Overrated

  I agree with a lot of this post. In particular, getting more precision in timelines is probably not going to help much with persuading most people, or in influencing most of the high-level strategic questions that Miles mentions. I also expect that it’s going to be hard to get much better predictions than we have now: much of the low-hanging fruit has been plucked. However, I’d personally find better timelines quite useful for prioritizing my technical research agenda problems to work on. I might be in a minority here, but I suspect not that small a one (say 25-50% of AI safety researchers).

  There’s two main ways timelines influence what I would want to work on. First, it directly changes the “deadline” I am working towards. If I thought the deadline was 5 years, I’d probably work on scaling up the most promising approaches we have now—warts and all. If I thought it was 10 years away, I’d try and make conceptual progress that could be scaled in the future. If it was 20 years away, I’d focus more on longer-term field building interventions: clarifying what the problems are, helping develop good community epistemics, mentoring people, etc. I do think what matters here is something like the log-deadline more than the deadline itself (5 vs 10 is very decision relevant, 20 vs 25 much less so) which we admittedly have a better sense of, although there’s still some considerable disagreement.

  The second way timelines are relevant is that my prediction on how AI is developed changes a lot conditioned on timelines. I think we should probably just try to forecast or analyze how-AI-is-developed directly—but timelines are perhaps easier to formalize. If timelines are less than 10 years I’d be confident we develop it within the current deep learning paradigm. More than that and possibilities open up a lot. So overall longer timelines would push me towards more theoretical work (that’s generally applicable across a range of paradigms) and taking bets on underdog areas of ML . There’s not much research into, say, how to align an AI built on top of a probabilistic programming language. I’d say that’s probably not a good use of resources right now—but if we had a confident prediction human-level AI was 50 years away, I might change my mind.

• AdamGleave 4 May 2022 23:00 UTC

  16 points

  0 ∶ 0

  on: In­for­ma­tion se­cu­rity con­sid­er­a­tions for AI and the long term future

  Thanks for the post! This seems like a clearly important and currently quite neglected area and I’d love to see more work on it.

  My current hot-take is that it seems viable to make AGI research labs a sufficiently hardened target that most actors cannot exploit them. But I don’t really see a path to preventing the most well-resourced state actors from at least exfiltrating source code. There’s just so many paths to this: getting insiders to defect, supply chain attacks, etc. Because of this I suspect it’ll be necessary to get major state actors to play ball by other mechanisms (e.g. international treaties, mutually assured destruction). I’m curious if you agree or are more optimistic on this point?

  I also want to note that espionage can reduce x-risk in some cases: e.g. actors may be less tempted to cut corners on safety if they have intelligence that their competitors are still far away from transformative AI. Similarly, it could be used as an (admittedly imperfect) mechanism for monitoring compliance with treaties or more informal agreements. I do still expect better infosec to be net-positive, though.

• AdamGleave 20 Apr 2022 17:06 UTC

  5 points

  0 ∶ 0

  in reply to: evhub’s comment on: Free-spend­ing EA might be a big prob­lem for op­tics and epistemics

  Making bets on new ambitious projects doesn’t seem necessarily at odds with frugality: you can still execute on them in a lean way, some things just really do take a big CapEx. Granted whether Google or any major tech company really does this is debatable, but I do think they tend to at least try to instill it, even if there is some inefficiency e.g. due to principal-agent problems.

• AdamGleave 14 Apr 2022 13:14 UTC

  35 points

  0 ∶ 0

  on: Free-spend­ing EA might be a big prob­lem for op­tics and epistemics

  Thanks for writing this post, this is an area I’ve also sometimes felt concerned about so it’s great to see some serious discussion.

  A related point that I haven’t seen called out explicitly is that monetary costs are often correlated with other more significant, but less visible, costs such as staff time. While I think the substantial longtermist funding overhang really does mean we should spend more money, I think it’s still very important that we scrutinize where that money is being spent. One example that I’ve seen crop up a few time is retreats or other events being organized at very short notice (e.g. less than two weeks). In most of these cases there’s not been a clear reason why it needs to happen right now, and can’t wait a month or so. There’s a monetary cost to doing things last minute (e.g. more expensive flights and hotel rooms) but the biggest cost is the event will be less effective than if the organizers and attendees had more time to plan for it.

  More generally I’m concerned that too much funding can have a detrimental effect on organisational culture. It’s often possible to make a problem temporarily go away just by throwing money at it. Sometimes that’s the right call (focus on core competencies), but sometimes it’s better off fixing the structural problem, before an organisation scales and it gets baked in. Anecdotally it seems like many of the world’s most successful companies do try to make frugality part of their culture, e.g. it’s one of Amazon’s leadership principles.

  In general, being inefficient at a small scale can still end up being very impactful if you work on the right problem. But I think to make a serious dent on the world’s problems, we’re likely going to need some mega-projects, spending billions of dollars with large headcount. Inefficiency at that scale is likely to result in project failure: oversight and incentives only get harder. So it seems critical that we continue to develop the ability in EA to execute on projects efficiently, even if in the short-term we might achieve more by neglecting that.

  I do feel a bit confused about what to do in practice to address these problems, and would love to see more thinking on it. For individual decisions, I’ve found figuring out what my time (in some context) is worth and sticking to it for time-money tradeoffs is helpful. In general I’d be suspicious if someone is always choosing to spend money when it saves time, or vice-versa. For funding decisions, these concerns are one of the reasons I lean towards keeping the bar for funding relatively high even if that means we can’t immediately deploy funding. I also support vetting people carefully to avoid incentivizing people pretending to be longtermists (or just having very bad epistemics).